ALMA OBSERVATIONS OF THE IRDC CLUMP G34.43+00.24 MM3: DNC/HNC RATIO

Figures 1(a)–(c) show the integrated intensity maps of DNC J = 3–2, N₂H⁺ J = 3–2, and HN¹³C J = 3–2. In these figures, the distribution is clearly different from molecule to molecule. The DNC emission (Figure 1(a)) peaks near the hot core (yellow star mark), and is extended toward the north of the hot core. A filamentary structure is also seen toward the south of the hot core. In Figure 1(b), the HN¹³C emission is very compact, and it peaks toward the hot core. No extended emission of HN¹³C is detected. On the other hand, the N₂H⁺ J = 3–2 emission is distributed throughout a relatively large area in this clump (Figure 1(c)). Near the hot core, the N₂H⁺ emission shows a clumpy structure. A filamentary structure is seen in N₂H⁺ toward the north of the hot core.

In Figure 1(d), we plot the CS J = 5–4 (Paper I) and DNC contours superposed on the N₂H⁺ color image. In the eastern part of the map, we can see a collimated outflow traced by CS. The CS emission seen in the southwestern part of the map should be outflows driven by embedded young stellar objects, as discussed in Paper I. In this figure, the N₂H⁺ emission is found to decrease toward the hot core and the outflow, while the DNC emission is detected near the hot core/outflow system.

Figure 2 shows the DNC and HN¹³C contours superposed on the CH₃OH J_K = 10₃–9₂ ${{A}^{-}}$ image (Paper I). Since the upper state energy of CH₃OH J_K = 10₃–9₂ A⁻ is 165 K, this line traces the hot core. In Figure 2, we can clearly see that the HN¹³C emission comes from the hot core and that the DNC peak is offset from the hot core.

Zoom In Zoom Out Reset image size

To investigate the relationship among the CS, N₂H⁺, and DNC distributions, we present velocity-channel maps in Figure 3. The CS emission appears over a wide velocity range, while the DNC and N₂H⁺ emission appears only in a relatively narrow velocity range. In the velocity range from 58 to 60 km s⁻¹, the N₂H⁺ emission is seen on the northern edge of the CS outflow lobe. Toward this region, the N₂H⁺ and CS distributions are clearly anti-correlated. This observational feature likely indicates that the outflow is interacting with the cold dense gas. This interpretation is also supported by detection of the class I CH₃OH maser toward the interacting region (Yanagida et al. 2014).

Zoom In Zoom Out Reset image size

In Figure 3, the DNC emission is seen near the interacting region at the velocity of 58–60 km s⁻¹. However, the DNC peak is clearly offset from the CS and N₂H⁺ peaks. Since the DNC peak is rather close to the hot core, DNC seems abundant in warm gas around the protostar. Furthermore, the offset of the DNC peak from the CS peak indicates that the DNC is not abundant in the interacting region.

In the velocities of 59.27–59.86 km s⁻¹, a filamentary structure in the DNC emission is seen toward the south of the hot core. Although the N₂H⁺ emission is detected in some parts of the filamentary structure of the DNC emission, most of the filament is only traced by the DNC emission.

As seen in Figures 1 and 2, the DNC emission is stronger than the HN¹³C emission in several positions inside the clump. Figure 4(b) shows the spectra of DNC and HN¹³C averaged within the 15$^{\prime\prime}$ × 15$^{\prime\prime}$ region around the phase center, where no primary beam corrections are applied. In this figure, the DNC emission is stronger than the HN¹³C emission. This is apparently inconsistent with the result of the single-dish observations of DNC J = 1–0 and HN¹³C J = 1–0 by Sakai et al. (2012), as shown in Figure 4(a). This discrepancy will be discussed later.

Zoom In Zoom Out Reset image size

In Figures 4(c) and (d), we present the DNC and HN¹³C spectra toward the DNC and HN¹³C peaks, respectively. Note that these spectra are not averaged. The DNC peak is about 08 offset from the HN¹³C peak. Toward the DNC peak, the DNC emission is stronger than the HN¹³C emission. The DNC spectrum shows a double peak, while the HN¹³C spectrum shows a single peak. Since the velocity of HN¹³C emission is coincident with the lower velocity component of the DNC emission, the double peak of the DNC emission corresponds to two different velocity components and is not a self-absorbed profile.

Toward the HN¹³C peak, the HN¹³C emission is stronger and broader than the DNC emission. In addition, the peak velocity of the HN¹³C line is offset from that of the DNC line. Since there are no other detectable molecular lines in this frequency range, except for the (CH₃)₂O lines seen around 75 km s⁻¹ (Figure 4(d)), the broad emission is only due to the HN¹³C emission. As mentioned above, the HN¹³C peak almost coincides with the hot core position. In Figure 4(d), we also plot the CH₃OH J_K = 10₂–${{9}_{3}}$ ${{A}^{-}}$ spectrum (Paper I). The velocity range of the HN¹³C emission is covered by that of the CH₃OH J_K = 10₂–${{9}_{3}}$ ${{A}^{-}}$ emission. Thus, the broad HN¹³C emission is likely tracing the hot core. Since the DNC emission is narrow toward the hot core position, it is likely that the DNC emission comes from a colder envelope rather than the hot core.

The emitting region of N₂H⁺ is found to be quite different from that of DNC. This suggests that these molecules trace different physical conditions. It is known that N₂H⁺ is less abundant in warm gas without CO depletion, because it is destroyed by reactions with CO. Thus, the weak N₂H⁺ emission near the hot core and the outflow indicates that the temperature is higher than the sublimation temperature of CO (∼20 K). In contrast, the N₂H⁺ emitting regions should be rather cold.

As for DNC, the main destruction mechanism is reactions with HCO⁺ or H₃⁺, and consequently, the destruction timescale of DNC is much longer than that of N₂H⁺, as mentioned in Section 1. Therefore, the DNC molecule is expected to survive at temperatures above 20 K for ∼10^3–4 yr. Since the N₂H⁺ emission is weak toward the DNC emitting regions, the DNC emission seems to trace relatively warm regions.

In addition, DNC may be formed on grain surfaces during the cold starless phase. The sublimation temperature of DNC and HNC is estimated to be about 80 K, if their binding energy is assumed to be the same as HCN (E/k = 4170 K) reported in Yamamoto et al. (1983). Thus, the abundances of HNC and DNC could increase in hot regions by sublimation from icy dust mantles. As shown in Figure 2, the DNC peak is offset from the hot core, and no heating source is probably associated with the DNC peak. The gas temperature of the DNC peak is therefore thought to be lower than that of the hot core. If the DNC abundance is enhanced due to sublimation, the HN¹³C abundance should be also enhanced. However, the HN¹³C emission is faint toward the DNC peak. Thus, it is most likely that the large DNC abundance toward the DNC peak does not originate from the sublimation of icy dust mantles. The temperature of the DNC emitting regions, except for the hot core, would be in a range from 20 to 80 K.

We derive the DNC/HNC abundance ratio from the spectra shown in Figures 4(b)–(d). We obtain the abundance ratio by taking the ratio between the DNC column density, N(DNC), and the HNC column density, N(HNC). The column densities were derived by assuming optically thin emission under local thermodynamic equilibrium conditions. For simplicity, we assume the same excitation temperature for DNC and HN¹³C. We calculate the column densities in the range previously defined: 20–80 K at the DNC peak and 20–130 K at the HN¹³C peak position.

In order to estimate N(HNC) from N(HN¹³C), Sakai et al. (2012) used the ¹²C/¹³C ratio of 62, which is estimated from the CO and H₂CO data by Wilson & Rood (1994). However, the HN¹²C/HN¹³C ratio may be higher than the ¹²CO/¹³CO ratio. According to the model calculations by Furuya et al. (2011), the HN¹²C/HN¹³C ratio could be higher by a factor of 2 than the ¹²CO/¹³CO ratio in dense (n > 10⁵ cm⁻³) regions. Hence, we assume the HN¹²C/HN¹³C ratio to be 60–80. The derived DNC/HNC abundance ratios are listed in Table 2.

Table 2.  DNC/HNC Abundance Ratio

Position	R.A.	Decl.	IDNC	${{I}_{{\rm H}{{{\rm N}}^{13}}C}}$	N(DNC)	N(DNC)/N(HNC)a
	(J2000.0)	(J2000.0)	(mJy beam−1 km s−1)	(mJy beam−1 km s−1)	1012 (cm−2)
DNC Peak	18 53 20.58	1 28 25.6	79 ± 3	24 ± 5	2.8–5.3b	0.03–0.08b
HN13C Peak	18 53 20.63	1 28 25.4	17 ± 4	130 ± 4	0.48–2.0c	0.001–0.004c
Averaged Spectra	...	...	1.65 ± 0.08	<0.3	0.058–0.11b	>0.06b

^aThe HN¹²C/HN¹³C ratio is assumed to be 60–80. ^bThe excitation temperature is assumed to be 20–80 K. ^cThe excitation temperature is assumed to be 20–130 K.

Download table as:  ASCIITypeset image

The DNC/HNC abundance ratio derived from the averaged spectra is higher than 0.06. This is much higher than that obtained from the single-dish observations (0.003 ± 0.001: Sakai et al. 2012). The DNC/HNC ratio toward the DNC peak (0.03–0.08) is also higher than that of the single-dish observation. These values are comparable to those observed in low-mass star forming regions (∼0.06; Sakai et al. 2012; Hirota et al. 2001, 2011).

Toward the HN¹³C peak, the derived DNC/HNC abundance ratio is 0.001–0.004. This ratio is comparable to that of the single-dish observation. However, the emission from the hot core is not the main component producing the emission detected with the single-dish telescope, because the hot core is compact and the spectral shape toward the HN¹³C peak is different from that of the single-dish observation (see Figure 4).

The DNC/HNC ratio is different between the single-dish and ALMA observations. One possibility for the weak HN¹³C emission is that the HN¹³C emission is so extended that its most part is resolved out in the ALMA observations. However, the HN¹³C emission should be detected toward the "compact" DNC emitting regions even in this case, as long as the DNC/HNC ratio is relatively low (∼0.003). Moreover, the critical density of the HN¹³C J = 3–2 emission is as high as ∼10⁷ cm⁻³, and it is unlikely that such high density regions are extended throughout the clump. Therefore, we do not think that the resolved-out problem is the main reason for the weak HN¹³C J = 3–2 emission in this ALMA observation.

The most probable explanation for the different DNC/HNC ratios is that the HN¹³C emission traces different density regions between the single-dish and ALMA observations. In the single-dish observations, we observed the J = 1–0 line. The critical density of HN¹³C and DNC J=1–0 (∼10⁵ cm⁻³) is lower than that of HN¹³C and DNC J = 3–2 (∼10⁷ cm⁻³). Therefore the single-dish observations preferentially trace the lower density envelopes, which may have a relatively low DNC/HNC ratio. Fontani et al. (2014) also pointed out that the DNC/HNC ratio derived from the J = 1–0 data traces the regions with a density of ∼10⁴ cm⁻³, based on the comparison between their single-dish observations and chemical model calculations. On the other hand, the ALMA observations mainly trace denser regions with higher deuterium fractionation ratios. The origin of the low DNC/HNC ratio in the diffuse regions will be discussed in the next section.

In Figure 5, we summarize our interpretation of the results in a schematic illustration. The clump consists of extended inter-core medium (∼10^4–5 cm⁻³) with embedded dense cores (∼10^6–7 cm⁻³). The DNC/HNC abundance ratio is relatively high in the densest regions, with the exception of the hot core. Since the filling factor of the dense regions is relatively low for the single dish J = 1–0 observation, the DNC/HNC ratio of the single dish observation is lower than that of the interferometer observations.The filling factor of the DNC emitting region within the 15$^{\prime\prime}$ × 15$^{\prime\prime}$ area is estimated to be about 0.1. Toward the hot core, the DNC/HNC ratio is rather low as compared with the nearby dense regions. The low DNC/HNC ratio in the hot core will be discussed in the next section.

Zoom In Zoom Out Reset image size

To investigate the behavior of the DNC/HNC abundance ratio in more detail, we carried out model calculations of gas-phase and grain-surface chemistry. We adopt the chemical reaction network of K. Furuya et al. (2015, in preparation); it is originally based on the reaction network of Garrod & Herbst (2006) and high-temperature gas-phase reaction network of Harada et al. (2010, 2012), and has been extended to include multideuterated species (Aikawa et al. 2012; Furuya et al. 2013) and nuclear spin states of H₂, H₃⁺, and their deuterated isotopologues (Hincelin et al. 2015, in preparation; Coutens et al. 2014). The rate coefficients of reactions between CH₃⁺ + H₂ and their deuterated isotologues are updated referring to Roueff et al. (2013). We adopt a three-phase model, which consists of gas, a chemically active ice mantle, and an inert ice mantle (Hasegawa & Herbst 1993); in the ice mantle, chemical reactions occur only in four layers from the surface (Vasyunin & Herbst 2013). Swapping between the surface active layer and innert ice mantle, i.e., the thermal diffusion of ice mantle, is included following Garrod (2013). The cosmic ray ionization rate is set to be 2.6 × 10⁻¹⁷ s⁻¹ (van der Tak & van Dishoeck 2000). Note that the binding energy of HNC used here (4170 K) is higher than that in (Sakai et al. 2012; 2050 K); the former is based on the measurement of vapor pressure of HCN (Yamamoto et al. 1983), while the latter value is adopted from Garrod & Herbst (2006). Thus, the sublimation temperature of HNC is ∼80 K in the current model, while that was ∼40K in the model of Sakai et al. (2012).

First, we assume a uniform cloud with constant temperature and constant density of 10⁴ or 10⁶ cm⁻³, and calculate the temporal variation of chemical compositions. Visual extinction was assumed to be 10 mag, so that photoreactions induced by the interstellar radiation field are unimportant. As an initial condition, all the elements are assumed to be in the form of neutral atom or atomic ion, depending on their ionization potentials, except for hydrogen and deuterium, which are in molecular form (H₂ and HD). The o/p ratio of H₂ affects the molecular D/H ratio, because the internal energy of ortho-H₂ acts as a reservoir of chemical energy which hampers the deuteration of H₃⁺. Recently, Brunken et al. (2014) measured the o/p ratio of H₂D⁺, and derived the o/p H₂ ratio to be ∼2 × 10⁻⁴ toward IRAS 16293-2422. It is not clear, however, what the optimal initial value of o/p ratio of H₂ is for our molecular cloud model; e.g., Flower et al. (2006) argued that the steady state value of the o/p ratio would not be attained before the protostellar collapse commences. Furuya et al. (2015, in preparation) recently calculated o/p ratio of H₂ in a molecular cloud formation. They showed that the o/p ratio of H₂ decreases to 10⁻³, when the column density of a cloud reaches ∼3 mag. We thus calculate models with the initial o/p ratio of 0.01 and 0.001.

Figures 6(a) and (b) show the model calculation results for the three temperature cases (10, 30, and 100 K) with a density of 10⁶ or 10⁴ cm⁻³, respectively. As shown in Figures 6(a) and (b), at a given time, the DNC/HNC abundance ratio is significantly higher in the high density model than in the low density model. This is mainly due to the longer chemical timescale in the lower density model. In addition, higher initial o/p ratio of H₂ molecules leads to lower D/H ratio, as seen in Figure 6. The o/p ratio may be higher in less dense regions (Pagani et al. 2011), which may contribute to relatively low deuterium fractionation in the low density regions. Thus, the low DNC/HNC ratio measured by using single-dish observations is likely due to low density gas traced by the J = 1–0 lines.

Zoom In Zoom Out Reset image size

In Figure 6(a), we can see that the DNC/HNC ratio observed toward the DNC peak is reproduced at t ∼ 10 ⁴–10⁵ yr at temperature of T < 30 K, while the DNC/HNC ratio at the HN¹³C peak is consistent with the model with T ∼ 100 K. The constant temperature model, however, would be unrealistic; the gas should have been cold initially, and then heated by the star formation. Thus, we next consider a sudden temperature rise from 10 K to a given temperature at the time of 3 × 10⁴ yr or 10⁵ yr, which simulates the birth of a protostar. Figures 7(b) and (c) show the temporal variation of the D/H ratio after the temperature-rise with a gas density of 10⁶ cm⁻³ and the initial o/p ratio of 0.01; t = 0 yr is defined as the time of the temperature-rise.

Zoom In Zoom Out Reset image size

In Figure 7(b), we can see that the DNC/HNC ratio does not decrease rapidly after the temperature rise, so that the DNC/HNC ratio toward the DNC peak could be reproduced with relatively high temperature (20–80 K) in the temperature-rise model. After the temperature-rise, the main formation reaction of DNC is dissociative recombination of HCND⁺, which is formed via DCO⁺ + HCN. The DCO⁺/HCO⁺ ratio at warm temperature is enhanced by HCO⁺ + D → DCO⁺ + H in our model calculations. Although deuterium enrichment of HNC via CH₂D⁺ (e.g., Roueff et al. 2007) works in the early time of our constant temperature models, those reactions are negligible in the warm phase of the temperature-rise models; most of the carbon is already locked into CO at the time of temperature-rise and CH₂D⁺ is not abundant.

We compare the observed DNC/HNC ratio at the HN¹³C peak with the model results at t ∼10³ yr, which is the dynamical timescale of the outflow (Paper I). In Figure 7, we can see that the DNC/HNC ratio depends on the ratio at the temperature-rise. If the temperature is raised at 1 × 10⁵ yr, the DNC/HNC ratio after the temperature rise is much higher than observed (Figure 7(b)). The observed value toward the HN¹³C peak is reproduced by the high temperature (>90 K) cases with the early temperature-rise (<3 × 10⁴ yr). Thus, the low DNC/HNC ratio toward the HN¹³C peak suggests that the temperature toward the HN¹³C peak have risen in relatively early time, so that the D/H ratio in the gas and ice (see below) at t = 0 is not too high.

In the model calculations, the low DNC/HNC ratio above 90 K originates from sublimation of grain surface molecules, whose DNC/HNC ratio at the cold phase (10 K) is lower than that in the gas phase, as seen in Figure 7(a). Figure 8(a) shows the DNC and HNC abundances in the ice mantle and in the gas phase for the constant temperature model (10 K), and Figure 8(b) for the temperature-rise models (from 10 to 60 or 100 K); the data plotted in Figures 8(a) and (b) are the same as those used in Figures 7(a) and (c), respectively. In Figure 8(b), we can see that both HNC and DNC abundances in the gas phase are enhanced by more than one order of magnitude after the temperature rises to 100 K, and remain abundant for >10³ yr. If both HNC and DNC were sublimated from dust grains efficiently in the hot core, both HNC and DNC abundances should be peaked toward the hot core in the clump. As mentioned before, this is not the case; the DNC peak is offset from the hot core.

Zoom In Zoom Out Reset image size

One possibility to solve this contradiction is a low beam filling factor of the hot core (HN¹³C peak); we proposed in Paper I that the hot core consists of a few unresolved substructures. Since DNC/HNC ratio in grain mantle is lower in that in the gas phase, DNC is not significantly enhanced toward the hot core unlike HN¹³C. This enhancement is further diluted by the beam filling factor in the observation. Thus, the enhancement of DNC is hardly recognizable in comparison with the overwhelming DNC emission around the nearby DNC peak. In order to investigate the low DNC abundance toward the hot core in more detail, higher angular resolution observations would be necessary. In addition, observations of DNC and HN¹³C toward other hot cores would be also necessary for a full understanding.

In this study, we have shown that the DNC/HNC abundance ratio is relatively high in the densest regions, while it is low in the diffuse envelope. Taking these results into account, we give another possible interpretation of the single-dish survey made by Sakai et al. (2012).

Sakai et al. (2012) reported that the DNC/HNC ratio in high-mass sources is lower than that in low-mass sources. One possible explanation is that the filling factor of the high density regions is different between observations of low-mass and high-mass objects. Since the low-mass sources are typically much closer to earth than the high-mass sources, the filling factor of the high-density regions may be higher in the low-mass sources than in the high-mass sources. In addition, it may also be possible that the DNC/HNC ratio of the high-mass sources is lower than that of the low-mass sources for a given density region (10⁴ cm⁻³) traced by the J = 1–0 lines of HNC and DNC, for instance, due to higher temperature. In order to investigate the difference between the low-mass and high-mass sources in more detail, multi-transition line observations with high angular resolution are therefore crucial.

Sakai et al. (2012) also reported that there is a diversity of the DNC/HNC ratio in the high-mass sources, where the DNC/HNC ratio is different even among the objects with similar evolutionary stages. Note that there is no correlation between the DNC/HNC ratio and the distance. A possible origin of this diversity is variation of the filling factor of high density regions; the objects with higher DNC/HNC ratio might contain more dense cores. Another possibility is that the DNC/HNC ratio of the densest regions differs from source to source. In this case, the ages of dense regions would be different from source to source. In either case, the diversity of the DNC/HNC ratio could reflect the difference in the history of cluster formations. Observations of the DNC/HNC ratio at a high spatial resolution have a high potential for elucidating the history of cluster formation.